Hi JitteryCoyote63 , can't you just use task.connect_configuration()
for that?
Hi SuccessfulKoala55 , super that’s what I was looking for
Hi, I have a configuration file that I read and connect to my training tasks. I cannot use config = task.get_parameters_as_dict()["General"]["param"]["nested"]["deep"]
because at some point, clearml just stringify the parameter (list/dict). Is there a parameter that can control that? I also have the problem that some nested parameters are int but converted to string by clearml
Hi JitteryCoyote63 , can't you just use task.connect_configuration()
for that?
Hi SuccessfulKoala55 , super that’s what I was looking for